We have used a monoclonal antibody directed against the C-terminus of the Drosophila inverted homeodomain to detect a nuclear protein in brain cells of Xenopus laevis embryos. We refer to this antigen as the Xenopus EN protein. The EN protein is localized at midneunda stage to a band of cells in the anterior portion of the neural plate, on each side of the neural groove. Later in development, the expression coincides with the boundary of the midbrain and hindbrain, and persists at least to the swimming tadpole stage. These properties make the EN protein an excellent molecular marker for anterior neural structures. In embryos where inductive interactions between mesodermal and ectodermal tissues have been perturbed, the expression of the EN protein is altered; in embryos that have been anterodorsalized by LiCl treatment, the region that expresses the EN protein is expanded, but still well organized. In ventralized UV-irradiated embryos, the absence of the protein is correlated with the absence of anterior neural structures. In extreme exogastrulae, where the contacts between head mesoderm and prospective neurectoderm are lost, the EN protein is not expressed.

Homeobox proteins are thought to act as sequencespecific transcription factors in diverse organisms (reviewed Levine & Hoey, 1988). The homeobox itself is a 180 bp sequence that is contained within the protein-coding region for genes whose protein products are invariably found in cell nuclei. Many homeobox genes are expressed in a restricted region of the vertebrate embryo. This property, combined with their proven role in regulating gene expression in invertebrate embryos, makes them good candidates for genes that bring about differentiation in the vertebrate embryo. Since these genes are expressed in a regionspecific way, they are good markers for early differentiation; in fact the region-specific expression of these genes within tissues demonstrates differentiation that has not been recognized by histological examination.

Though similar in sequence, the homeoboxes have been divided into groups based on the extent of sequence similarity within as well as outside the homeobox. The engrailed homeobox defines one such class which contains the engrailed and inverted genes from Drosophila as well as the En-1 and En-2 genes in the mouse (reviewed in Joyner & Martin, 1987; Barad et al. 1988). In the fruit fly, the two members of this class engrailed (en) and inverted (inv) are involved in segmentation (Kornberg, 1981; Coleman et al. 1987). Both the en and inv RNAs (Kornberg et al. 1985; Coleman et al. 191) and proteins (DiNardo et al. 1985) are localized in the posterior compartment of each segment and are involved in determination of cell fate of the posterior compartment (Kornberg, 1981). Two mouse genes, En-1 and En-2, have been characterized in the mouse by virtue of their sequence similarity to the en homeobox (Joyner & Martin, 1987). The transcripts of both genes have been localized by in situ hybridization (Davis et al. 1988; Davis & Joyner, 1988; Davidson et al. 1988). En-2 is expressed in the developing midbrain at the junction with the hindbrain while En-1 is expressed in all cells expressing En-2 as well as cells in a restricted segment of the neural tube, dermatomes of tail somites, developing vertebrae, facial mesenchyme and limb-bud ectoderm (Davidson et al. 1988; Davis et al. 1988; Davis & Joyner, 1988).

Recently Patel et al. (1989) have shown that an antibody (4D9) that recognizes the C-terminus of both en and znv’homeodomains detects an antigen that is expressed in embryos of diverse species including chicken, zebrafish and Xenopus. Such conservation argues further for the functional importance of this class of genes during embryogenesis.

Following the preliminary identification of the reactive antigen in Xenopus by Patel et al. (1989), we pursued the question of when and where the gene encoding the engrailed antigen is expressed. We have also analyzed the cell interactions necessary to express the antigen. We show that the antibody recognizes a protein expressed at the junction of the midbrain and hindbrain. We have found that the antibody reveals differentiation within the brain well before neural tube closure. Thus the en antibody will be a useful tool in studying the regional specificity of neural induction. Although molecular markers of neural induction are available, they are mostly expressed throughout neural tissue, and only one region-specific marker is currently available, marking spinal cord differentiation (XlHboxó-, Sharpe et al. 1987). To begin an analysis of the cell interactions necessary to turn on expression of the antigen, we have used UV irradiation, LiCl treatment and hypertonic medium to produce ventralized, dorsalized or exogastrula embryos (Scharf & Gerhart, 1980, 1983; Kao et al. 1986; Kao & Elinson, 1988; Holtfreter, 1933). The presence of the antigen in such experimentally perturbed embryos correlates well with the induction of anterior neural structures.

Since the antibody staining we see in Xenopus corresponds closely to the spatial distribution of transcripts from the En-2 gene in the mouse, we suspect that the antibody recognizes a Xenopus homolog to En-2. For convenience we refer to the protein recognized by the antibody as the Xenopus EN protein. Direct proof that the antibody recognizes the product of a gene homologous to En-2 must await the isolation and sequencing of the Xenopus gene.

Embryos

For all the studies described, albino Xenopus laevis were used. These albinos were obtained from the Berkeley colony maintained by the laboratory of J. C. Gerhart. Ovulation of females, in vitro fertilization, LiCl treatment, UV irradiation and exogastrulae were carried out as described by Condie & Harland (1987). Stages according to Nieuwkoop & Faber (1967).

The antibody

The 4D9 antibody is a monoclonal IgG kindly provided to us by N. Patel and C. Goodman, UC Berkeley. The antibody was originally raised against the Drosophila inverted protein. The epitope recognized by this antibody has been mapped to the C-terminal region of the homeobox sequence, and is contained within a 15 amino acid peptide with the following sequence: SSELGLNEAQIKIWF. The glycine residue at position 5 seems to be a crucial amino acid for the recognition by the monoclonal antibody. A substitution of this glycine to a serine completely abolishes recognition of the antibody (Patel et al. 1989).

Immunohistochemistry

The protocol of Patel et al. (1989) was adapted to Xenopus embryos as follows. All embryos were fixed in MEMFA (0-1m-Mops pH 7·4, EGTA 2mm, MgSO4 liriM, Formaldehyde 3·7%) for l-2h at room temperature. The vitelline membrane of early embryos was removed manually during fixation. Embryos were then stored at −20°C in methanol until staining. Staining was carried out in 5 ml screw-cap glass vials. Embryos were rehydrated gradually with PBS (NaH2PO4 1·85 mm, Na2HPO4 8·5 mm, NaCl 175 mm, pH adjusted to 7·4) over a period of 10 to 15 min. Non-specific protein-binding sites on embryos were blocked by incubation at room temperature in 5 ml PBT (PBS+2 mg ml-1 BSA+0·1% Triton X-100) with continuous rotation (end over end, 12 revs min-1) for 15 min. This solution was replaced with 0·5 ml PBT+10% goat serum (Gibco heat inactivated at 56°C for 30 min) to block non-specific immunoglobulin-binding sites for 1 h at room temperature with rocking on a nutator (Clay Adams). The appropriate dilution of the primary antibody (1:1 for the 4D9 cell culture supernatant and 1:1000 for ascites fluid containing monoclonal antibody) was added to 0·5ml fresh PBT+10% goat serum and incubated overnight at 4 °C on a nutator. Embryos were washed the next day with 3 to 5 changes, 2h each, of PBT at room temperature. The secondary antibody was a goat anti-mouse IgG coupled to HRP (Biorad). The incubation of the embryos with secondary antibody was also done in PBT+10% goat serum at a dilution of 1:1000, overnight at 4°C with gentle shaking. Embryos were then washed as described above and stained by adding 1ml of diaminobenzidine (DAB; Polyscience, resuspended in PBT as 0-5 mg tnl-1 solution and filtered) with vigorous swirling at 4°C for 15 min. This is to allow the DAB to penetrate the embryo before the hydrogen peroxide is added so that the staining is homogeneous inside and outside the embryos. Finally the color reaction is visualized by adding 1 μl of 30 % H2O2 diluted 1:1 in PBT to each tube with continuous vigorous swirling at 4°C for 1 to 2h. Since there is extremely low background with this antibody, we have used NiCl2 in some cases to enhance the HRP stain (Figs 1A, IB, 3A, 5, 6 and 8). This is done by adding, simultaneously with the DAB, NiCl2 to a final concentration of 0·04%. At the end of staining, embryos were dehydrated with two washes of methanol and made transparent by incubation with a mixture of benzyl benzoate (Mallincrodt, OR grade) and benzyl alcohol (Fisher) at 2:1 ratio, respectively, prior to examination by light microscopy, a procedure devised by Andrew Murray (personal communication).

Fig. 1.

The onset of expression of the EN protein in wholemount embryos. (A) Anterior view of early neurula embryo stage 14 (Nieuwkoop & Faber, 1967). (B) Anterior view of early neurula stage 15. The arrows indicate the position of the neural groove.

Fig. 1.

The onset of expression of the EN protein in wholemount embryos. (A) Anterior view of early neurula embryo stage 14 (Nieuwkoop & Faber, 1967). (B) Anterior view of early neurula stage 15. The arrows indicate the position of the neural groove.

A disadvantage of albino embryos is that it is more difficult to stage them and to see enough detail to carry out surgical manipulations. This disadvantage can be alleviated by vital staining the whole embryo with Nile blue. This is done by incubating the albino embryos in a 0·01 % solution of Nile blue in 66mm-sodium phosphate pH 7-8 plus 5% Ficoll for 10 min at room temperature (T. Doniach personal communication). Embryos are subsequently washed with 33% MMR (1XMR= lOOmm-NaCl, l·8mm-KCl, l·0mm-MgCl2, 2·0mm-CaCl2, 50/4ml-1 gentamycin, buffered to pH6-9 with 5mm-Hepes) and kept in the same solution. This allows easy identification of pigment lines at the blastopore, neural fold etc. The Nile blue is removed during methanol washing.

Embryos to be sectioned were transferred from the benzyl benzoate/benzyl alcohol to methanol for 10 min and subsequently embedded in paraplast. 10 μm sections were mounted in benzyl benzoate/benzyl alcohol.

Expression in normal embryos

The onset of expression

We have used albino embryos to aid in the detection of a low abundance antigen in whole-mount immunohistochemistry. We first detect the EN protein in early neurula embryos (stage 14) as shown in Fig. 1. The neural plate has begun to form and the segregation of the neural area from the surrounding epidermis is visible on the dorsal side of the embryo. In albino embryos stained with Nile blue, the neural folds are visible at this stage but they are still well separated from one another. Sections of such embryos show that the notochord and somitic mesoderm are discernible but individual somites have not yet begun to segregate (data not shown). The EN protein in frogs therefore seems to be expressed at an equivalent stage to the murine En-2 RNA which is detected by in situ hybridization at day 8-0 of gestation (Davis et al. 1988; Davis & Joyner, 1988; Davidson et al. 1988), though in the mouse five somites have segregated at the time of detection of the En-2 RNA.

Spatial distribution of the EN protein

The protein is localized to the nuclei of cells that are arranged as a stripe across the anterior neural plate, with an interruption at the center fine (the neural groove) as shown in Fig. 1A and IB. Serial sagittal and frontal sections across the embryos show that staining occurs in both the superficial and deep layers of neural plate ectoderm and that approximately 600 nuclei are expressing the EN protein (data not shown). The embryo at this stage contains DNA for approximately 60000 cells (Dawid, 1965). Therefore around 1−2% of the cells of a neurula embryo are expressing the protein.

Following this stage, the neural plate continues to fold into the neural tube. All the cells expressing the protein follow this folding process and ultimately form an almost complete ring around the neural tube. This pattern of expression is maintained with no detectable change into the tailbud stage (Figs 2A and 3A). Sections of tailbud tadpoles show that the staining coincides with the boundary of the mesencephalon and the rhombencephalon, two of the three brain vesicles formed in the neural tube by the évagination of the brain wall. In the early tailbud stage, cells expressing the EN protein are concentrated within a zone with sharp anterior and posterior boundaries which spans the posterior wall of the midbrain and the anterior wall of the hindbrain (Figs 2B and 3C). As shown in anterior view in Fig. 3B, staining is restricted to the sides of the neural tube and is discontinuous at the dorsal and ventral extremities.

Fig. 2.

Localization of the EN protein in whole-mount tailbud (stage 28). (A) The EN protein is expressed only in the brain. (B) Higher magnification view in the plane of the vesicles of the brain.

Fig. 2.

Localization of the EN protein in whole-mount tailbud (stage 28). (A) The EN protein is expressed only in the brain. (B) Higher magnification view in the plane of the vesicles of the brain.

Fig. 3.

Expression of the EN protein in the swimming tadpole (stage 39). (A) Localization in the whole animal behind the eye but anterior to the otic vesicle. (B) Optical section of whole mount in anterior view (equivalent to transverse section), showing that the protein is present in the lateral walls of the brain only. (C) Optical section, side view of the same embryo (sagittal section).

Fig. 3.

Expression of the EN protein in the swimming tadpole (stage 39). (A) Localization in the whole animal behind the eye but anterior to the otic vesicle. (B) Optical section of whole mount in anterior view (equivalent to transverse section), showing that the protein is present in the lateral walls of the brain only. (C) Optical section, side view of the same embryo (sagittal section).

An interesting aspect of EN expression is that in the swimming tadpole (stage 36 to 42), the boundaries of EN expression become less sharp (Fig. 4). We find that a few cell nuclei outside the original band, within both the hindbrain and the midbrain, contain the EN protein. We do not know if this new distribution (Fig. 4) is a reflection of cell migration or new expression of EN. Finally in the feeding tadpole (stage 47), we find that the signal is reduced in intensity (Fig. 5) and mostly restricted to the boundary between the midbrain and hindbrain. The reduction in signal is likely due to a combination of factors; first a reduction in the quantity of EN protein and second a slight decrease in the efficiency of the whole-mount immunohistochemistry at late stages. The latter possibility is based on immunohistochemical staining of β-galactosidase expressed from injected plasmids (Hemmati Brivanlou unpublished results); in these experiments the efficiency of detection of β-gal in late stages was reduced relative to earlier stages. We cannot detect the more posterior cells which were seen in Fig. 4; however, we detect the two more anterior patches of cells expressing the protein in the floor of the mesencephalon (Fig. 4). Each patch contains approximately 30 to 40 cells (data not shown).

Fig. 4.

Optical section at the midbrain hindbrain boundary of the brain (stage 42). In addition to the expression at the boundary of the mid- and hindbrain, EN expression is also detected in the cells of the hindbrain (h) and the floor of the midbrain (m, see arrow).

Fig. 4.

Optical section at the midbrain hindbrain boundary of the brain (stage 42). In addition to the expression at the boundary of the mid- and hindbrain, EN expression is also detected in the cells of the hindbrain (h) and the floor of the midbrain (m, see arrow).

Fig. 5.

EN expression in the feeding tadpole stage 47 (arrow). The staining in the blood vessels is nonspecific and due to the peroxidase activity associated with heme in red blood cells.

Fig. 5.

EN expression in the feeding tadpole stage 47 (arrow). The staining in the blood vessels is nonspecific and due to the peroxidase activity associated with heme in red blood cells.

At all times we find the EN protein in cell nuclei, as shown in the section of anterior neural plate in Fig. 6.

Fig. 6.

Transverse section of a midneurula stained with the 4D9 antibody. The staining is restricted to the nuclei (see arrow). Bar, 10μm.

Fig. 6.

Transverse section of a midneurula stained with the 4D9 antibody. The staining is restricted to the nuclei (see arrow). Bar, 10μm.

Expression of EN in experimentally perturbed embryos LiCl-treated embryos

Since EN expression is very tightly localized to anterior neural tissues, it is of interest to examine EN expression in embryos where anterior and posterior cell identity has been altered. It has been demonstrated previously (Kao et al. 1986; Breckenridge et al. 1987; Cooke & Smith, 1988) that when early Xenopus laevis embryos (stage 6) are exposed to Li+ they develop exaggerated head and other dorsoanterior structures. This change in cell fates correlates with the suppression of expression of the Xhox-36 gene whose transcripts are normally found in posterior neural tissue as well as posterior mesoderm and epidermis (Condie & Harland, 1987; B. Condie unpublished results). In the most extreme cases, Li+-treated embryos exhibit radial symmetry with expanded head-like structures in which the eye and cement gland extend radially around the embryo (Kao & Elinson, 1988). Breckenridge et al. (1987) found that the CNS was completely disorganized and suggested that this was the result of difficulties during neurulation caused by an expanded neural plate. We have used LiCl-treated embryos to ask whether the anterior neural plate is indeed expanded during neurulation and if so how the EN protein is affected by the increase in dorso anterior cell specification. As expected the EN protein displayed a different pattern of expression in these embryos. First, the number of nuclei expressing the protein increased by a factor of about five (based on counting nuclei in sectioned embryos, data not shown) compared to the control embryo (Fig. 7A). Second, the band of cells expressing the protein extends most of the way around the embryo (Fig. 7B). The stage at which EN protein expression is detected remains unchanged compared to normal embryos whereas the spatial regulation is reprogrammed; the result is a band of cells expressing the protein over a much larger area than normal, confirming that this region of the neural plate is enormously expanded by the LiCl treatment.

Fig. 7.

Reprogramming of the EN expression in LiCl-treated embryos. (A) Untreated stage-18 control embryo. (B) Embryo previously treated with LiCl and fixed at the same age as the embryo in A. Arrows delimit the radial band of nuclei staining with the antibody. (C) Comparison of expression of the EN protein in normal (top) versus LiCl treated embryo (bottom), at late stage of embryogenesis. The EN protein occupies the subapical position of the anterodorsalized embryo.

Fig. 7.

Reprogramming of the EN expression in LiCl-treated embryos. (A) Untreated stage-18 control embryo. (B) Embryo previously treated with LiCl and fixed at the same age as the embryo in A. Arrows delimit the radial band of nuclei staining with the antibody. (C) Comparison of expression of the EN protein in normal (top) versus LiCl treated embryo (bottom), at late stage of embryogenesis. The EN protein occupies the subapical position of the anterodorsalized embryo.

This result complements that obtained with Xhox-36 (Condie & Harland, 1987) and is consistent with Li+ inducing an anteroposterior respecification of cell fate. Perhaps surprisingly, differentiation within this expanded region of the neural plate remains well ordered. The cells expressing the EN protein are still found in a narrow band rather than being expressed at random over a larger area of the prospective neural plate. The disorganization of neural structures seen in later stages by Breckenridge et al. (1987) may be a subsequent event, or a result of the prolonged treat-ment with LiCl used by these workers. In contrast to this disorganization, we find that when control embryos have reached the tadpole stage Li+-treated embryos still express the EN protein in a well-ordered ring at one end of the embryo (Fig. 7C). Since the EN protein is restricted to the boundary between midbrain and hindbrain, our observations are consistent with those of Cooke & Smith (1988) who examined embryos displaying an extreme LiCl syndrome and found that the nervous tissue occupying an apical position near the blastopore lacked any forebrain characteristics.

UV-treated embryos

Whereas LiCl embryos were predicted to have excessive expression of the EN protein, we expected that ventralized embryos would display a reduction of expression of the protein. Such reduction of expression would allow us to correlate expression of the EN protein with the presence of various tissues in the anterior-posterior axis.

UV treatment of Xenopus embryos reproducibly results in embryos lacking obvious dorsal structures such as notochord and neural plate (Malacinski et al. 1975; Scharf & Gerhart, 1980, 1983). In the most extreme case, the embryo has no externally visible somites, little or no tail fin and is almost radially symmetrical about the former animal-vegetal axis (grade 5 UV embryo, Scharf & Gerhart, 1980); in such embryos, no EN protein is detectable (Fig. 8A). The same result is observed in acephalic embryos lacking head structures (grade 4). In less severe cases where the rudiments of anterior structures are present (grade 1 and 2, microcéphalie and cyclopic embryos), we detect a small number of cells expressing the EN protein anterior to the otic vesicle (Fig. 8B).

Fig. 8.

Expression of EN protein in UV-treated ventralized embryo (age equivalent to swimming tadpole stage 42).

(A) The EN protein is not present in the most extreme case of ventralized embryo (grade 5). (B) High-magnification view of the brain of a synophthalmic (grade 2) embryo. The expression of the EN protein is reduced in this embryo (compare this profile of expression with the one depicted in Fig. 4). m, midbrain; h, hindbrain.

Fig. 8.

Expression of EN protein in UV-treated ventralized embryo (age equivalent to swimming tadpole stage 42).

(A) The EN protein is not present in the most extreme case of ventralized embryo (grade 5). (B) High-magnification view of the brain of a synophthalmic (grade 2) embryo. The expression of the EN protein is reduced in this embryo (compare this profile of expression with the one depicted in Fig. 4). m, midbrain; h, hindbrain.

Since the mesoderm in extreme UV embryos is ventral in nature (Cooke & Smith, 1987), and ventral mesoderm is not capable of inducing neural structures, the absence of EN expression was expected. In less severe cases, where rudiments of dorsal mesoderm are left, the expression of the EN protein correlates with the presence of midbrain structures.

Exogastrulae

The presence of mesoderm is essential for the proper determination of the neuroectoderm (for review, see Gurdon, 1987). Holtfreter (1933), working with newt embryos, showed that culturing embryos in hypertonic solution prevented the normal invagination of prospective mesoderm during gastrulation. The prospective mesoderm fails to move inward and beneath the ectoderm but evaginates and extends outside the ectoderm. The resulting exogastrulae develop some normal mesodermal structures (somites, notochord, head mesoderm), but fail to develop an organized nervous system (Holtfreter, 1933). However, the contact between ectoderm and mesoderm has been shown to be sufficient to induce a surprisingly large amount of neural tissue as demonstrated by the detection of the neural marker N-CAM (Kintner & Melton, 1987). In contrast, transcripts of the homeobox-containing gene XlHboxó (Sharp et al. 1987) which are normally expressed in the posterior region of the CNS are only detected at trace levels in extreme exogastrulae. A different homeobox-containing gene, Xhox-36, whose transcripts are found in posterior mesodermal and ectodermal tissues (Condie & Harland, 1987), is expressed at normal levels in extreme exogastrulae. In view of these observations, it was of interest to ask whether the nervous tissue formed in Xenopus exogastrulae might express EN as an anterior neural marker. We find that unlike N-CAM and Xhox-36 (Kintner & Melton, 1987; Condie & Harland, 1987) the EN protein is not present in the most extreme cases of exogastrulae (Fig. 9D). In contrast incomplete exogastrulae, where a considerable amount of prospective anterior mesoderm has involuted, do express the EN protein (Fig. 9B and 9C).

Fig. 9.

Expression of EN in partial and complete exogastrulae. (A) Control embryo without exposure to high salt (late neurula stage 20). (B and C) Incomplete exogastrulae showing expression of EN protein (arrows). (D) Complete exogastrula. The ectoderm is the folded vesiculated structure at the right.

Fig. 9.

Expression of EN in partial and complete exogastrulae. (A) Control embryo without exposure to high salt (late neurula stage 20). (B and C) Incomplete exogastrulae showing expression of EN protein (arrows). (D) Complete exogastrula. The ectoderm is the folded vesiculated structure at the right.

Taken together the results obtained from the perturbed embryos demonstrate that EN is induced in the neuroectoderm by the underlying mesoderm. This induction requires a more extensive contact than that maintained in extreme exogastrulae.

Normal expression of the EN protein

We have used an antibody raised against the Drosophila invected protein to identify an antigen expressed exclusively in the brain of Xenopus embryos. The similarity of homeobox sequences has meant that a large variety of homeobox-containing genes can be isolated by virtue of DNA sequence similarity; in contrast, the monoclonal antibody raised against inverted recognizes only a small subset of horneo-domains, presumably because of the requirement for strong conservation of a C-terminal portion of the homeobox, (Patel et al. 1989). In the case of this antibody (4D9), the antigen most commonly recognized in vertebrates is expressed at the boundary of the mid- and hindbrain (Patel et al. 1989). This expression is strikingly similar to the expression of the En-2 gene in the mouse and we believe that the expression we see in Xenopus probably arises from the gene homologous to En-2, though this remains to be proved. The antibody does not detect cells in the spinal cord, where the mouse En-1 gene is expressed (Davidson et al. 1988; Davis & Joyner, 1988). Thus, if there is a gene homologous to En-1 in Xenopus the antibody may not detect its product.

The antibody used in this study reacts with cell nuclei in the anterior neural plate as early as midneurula (Nieuwkoop & Faber, 1967, stage 14-15) showing that differentiation within the neural plate starts shortly after gastrulation even though there is no obvious morphological differentiation of this region of the prospective brain (Fig. 1A and IB). At this stage, the neural tube is completely open and no somites have segregated in the underlying mesodermal mantle. Thus the timing of expression is similar or slightly earlier than the start of expression of En-2 detected by in situ hybridization in the mouse. Expression in the mouse is detected on day 8-0 of gestation, at which time the neural tube is not yet closed and 5 somites have segregated (Davis et al. 1988). In Xenopus a marker of posterior neural plate, XlHboxó, is also detected after gastrulation, so that regional patterning of the entire neural plate may be complete in the early neurula. This early differentiation is consistent with patterning in urodele amphibians, where transplantation experiments show that regions of the neural plate are committed by midneurula (Kallen et al. 1965).

The later expression of the EN protein in Xenopus is restricted to a narrower band of cells than in the other vertebrates that have been examined (Patel et al. 1989), but occupies a similar position between the mesencephalon and rhombencephalon (mid- and hindbrain; Fig. 2B). The highly localized expression is consistent with the proposal that the protein is involved in the establishment of spatial boundaries within the brain (Davis et al. 1988, Davidson et al. 1988). By swimming tadpole (stage 39) a restricted number of cells posterior and anterior to the main band of expression are also expressing the EN protein. These may arise either by migration of cells out of the main band of expressing cells or they may represent different cells that express the gene for the first time at the later stage.

It is particularly useful that in Xenopus the antigen is only detected in the brain since it allows use of the antibody as a molecular marker for anterior nervous system. It is of critical importance that an array of neuronal markers be made available for different regions of the nervous system if patterning of the nervous system during neural induction is to be understood (Sharpe et al. 1987; Richter et al. 1988). So far most of the neural markers available for Xenopus are general markers. These include nerve cell adhesion molecule N-CAM (Jacobson & Rutishauser, 1986; Kintner & Melton, 1987; Levi et al. 1987), the nerve-cell-specific intermediate filament NF-M (Sharpe, 1988) and a series of cloned markers (Richter et al. 1988). It is also possible to use markers for epidermis which are turned off during neural induction such as cytokeratin (Jamrich et al. 1987) or Epi-1 (London et al. 1988). The only regional marker so far isolated is the XlHboxó gene which is expressed in posterior nerve cord and for which nucleic acid probes, but not antibodies, have been used. The availability of the engrailed monoclonal antibody will allow questions to be asked about the competence of different regions of the prospective neural plate to express anterior markers; in addition, we will be able to further test the importance of the mesoderm in patterning the nervous system. The single cell resolution offered by the antibody will allow detailed examination of small tissue expiants and combinations, thus allowing exploitation of the experimental embryology of the amphibian embryo in which neural induction was first discovered (Spemann & Mangold, 1924; Mangold, 1933), and in which the properties of both mesodermal and neural induction are under intense scrutiny.

Expression of the EN protein in experimentally perturbed embryos

In this paper, we have also addressed the question of what happens to the neural plate in dorsoanteriorized embryos and ventralized embryos. Dorsoanteriorized embryos can be induced by LiCl treatment during cleavage (Kao & Elinson, 1988). Posterior structures are suppressed in these embryos as shown by the reduction of Xhox-36 expression (Condie & Harland, 1987). We found that, in the early neurula stage, the pattern of expression of the EN protein in LiCl embryos was well organized into a single band of cells, but the band of cells was greatly extended compared to that in control embryos and, in extreme cases, extended radially around the embryo (Fig. 7A and 7B). Different regimes of LiCl treatment result in different degrees of order in the LiCl-treated embryos. Severe treatment still allows neural tissue to be formed, though in an apparently disorganized fashion (Breckenridge et al. 1987); in contrast, milder treatment can result in well-organized tissues in the extreme dorsoanterior embryos with neural tissue extending radially around one end (Cooke & Smith, 1988). Cooke and Smith inferred that the neural tissue was of mid- and hindbrain character because it was close to retinal and otic vesicle structures. Using the antibody against engrailed, we have shown that much of the neural tissue formed is indeed characteristic of the mesencephalon-rhombencephalon boundary, since it expresses the EN protein (Fig. 7C).

The pattern of expression of the EN protein in ventralized embryos is a reflection of the amount of anterior pattern remaining. The antigen is absent from extreme aneural embryos (grade 5; Scharf & Gerhart, 1983), but is present in a few cells of intermediate embryos which retain some eye tissue (Fig. 8). The expectation that neural tissue of anterior character is only formed in response to inductive signals from anterior mesoderm is corroborated in exogastrula embryos (Fig. 9). In extreme exogastrulae, the limited contact between ectoderm and mesoderm is sufficient to induce some neural tissue, as shown by expression of N-CAM (Kintner & Melton, 1987); however, we never detect the EN protein in such embryos. The posterior neural marker XlHboxó shows only traces of expression in these embryos (Sharpe et al. 1987) though the marker of posterior mesodermal and ectodermal tissues, Xhox-36, is expressed at normal levels (Condie & Harland, 1987). Intermediate stages of exogastrulae, in which the prechordal mesoderm and varying amounts of chordal mesoderm have involuted, express XlHboxó (Sharpe et al. 1987) and also express some EN protein (Fig. 9). These results are consistent with the notion that regional contact between mesoderm and ectoderm is required for at least some aspect of patterning within the nervous system. The EN protein should provide a useful marker for experiments to elucidate the molecular cues that elicit patterning in the nervous system.

We are indebted to Nipam Patel and Corey Goodman (UC Berkeley) for the gift of the 4D9 cell fine. We thank John Gerhart, Carey Phillips, Brian Condie, Chris Kintner and Nipam Patel for advice and critical reading of the manuscript. This work was supported by the NIH.

Barad
,
M.
,
Jack
,
T.
,
Chadwick
,
R.
&
Mcginnis
,
W.
(
1988
).
A novel, tissue specific, Drosophila homeobox gene
EMBO J.
7
,
2151
2161
.
Breckenridge
,
L. J.
,
Warren
,
R. L.
&
Warner
,
A. E.
(
1987
).
Lithium inhibits morphogenesis of the nervous system but not neuronal differentiation in Xenopus laevis
.
Development
99
,
353
370
.
COLEMAN
,
K. G.
,
Poole
,
S. J.
,
Weir
,
M. P.
,
Soeller
,
W. C.
&
Kornberg
,
T.
(
1987
).
The tnvected gene of Drosophila’, sequence analysis and expression studies reveal a close kinship to the engrailed gene
.
Genes and Dev.
1
,
19
28
.
Condie
,
G. C.
&
Harland
,
R. M.
(
1987
).
Posterior expression of a homeobox gene in early Xenopus embryos
.
Development
101
,
93
105
.
Cooke
,
J.
&
Smith
,
E. J.
(
1988
).
The restrictive effect of early exposure to lithium upon body patterning Xenopus development, studied by quantitative anatomy and immunofluorescence
.
Development
102
,
85
99
.
Cooke
,
J.
&
Smith
,
J. C.
(
1987
).
The midblastula cell cycle transition and the character of the mesoderm in u.v.-induced nonaxial Xenopus development
.
Development
99
,
197
210
.
Davidson
,
D.
,
Graham
,
E.
,
Sime
,
C.
&
Hill
,
R.
(
1988
).
A gene with sequence similarity to Drosophila engrailed is expressed during the development of the neural tube and vertebrae in the mouse
.
Development
104
,
305
316
.
Davis
,
C. A.
,
Noble-Topham
,
S. E.
,
Rossant
,
J.
&
Joyner
,
A. L.
(
1988
).
Expression of the horneo box-containing gene EN-2 delineates a specific region of the developing mouse brain
.
Genes Dev
.
2
,
361
371
.
Davis
,
C. A.
&
Joyner
,
A. L.
(
1988
).
Expression patterns of the horneo box-containing genes En-1 and En-2 and the protooncogene int-1 diverge during mouse development
.
Genes Dev
.
2
,
1736
1744
.
Dawid
,
I. B.
(
1965
).
Deoxyribonucleic acid in amphibian eggs
.
J. molec. Biol
.
12
,
589
.
Dinardo
,
S.
,
Kuner
,
J. M.
,
Theis
,
J.
&
O’farrell
,
P.
(
1985
).
Development of embryonic pattern in D. melanogaster as revealed by accumulation of the nuclear engrailed protein. 1985
.
Cell
43
,
59
69
.
Gurdon
,
J. B.
(
1987
).
Embryonic induction-molecular prospects
.
Development
99
,
285
306
.
Holtfreter
,
S.
(
1933
).
Die total exogastrulation, eine selbstabloesung des ektoderms von entomesoderm. Wilhelm Roux Arch. EntwMech. Org
.
129
,
669
793
.
Jacobson
,
M.
&
Rutishauser
,
U.
(
1986
).
Induction of neural cell adhesion molecule (N-CAM) in Xenopus embryos
.
Devi Biol
.
116
,
524
531
.
Jamrich
,
M.
,
Sargent
,
T. D.
&
Dawid
,
I. B.
(
1987
).
Cell-type-specific expression of epidermal cytokeratin genes during gastrulation of Xenopus laevis
.
Genes Dev
.
1
,
124
132
.
Joyner
,
A. L.
&
Martin
,
G. R.
(
1987
).
En-1 and En-2, two mouse genes with sequence homology to the Drosophila engrailed gene: expression during embryogenesis
.
Genes Dev
.
1
,
29
38
.
Kallen
, (
1965
).
Organogenesis
(ed.
R. L.
Dehaan
and
H.
Ursprung
).
New York, Holt, Rinehart and Winston
.
Kao
,
K. R.
&
Elinson
,
R. P.
(
1988
).
The entire mesodermal mantle behaves as Spemann’s organizer in dorso-anterior enhanced Xenopus laevis embryos
.
Devi Biol
.
127
,
64
77
.
Kao
,
K. R.
,
Masui
,
Y.
&
Elinson
,
R. P.
(
1986
).
Lithium-induced respecification of pattern in Xenopus laevis embryos
.
Nature, Land
.
322
,
371
373
.
Kintner
,
C. R.
&
Melton
,
D. A.
(
1987
).
Expression of Xenopus N-CAM RNA in ectoderm is an early response to neural induction
.
Development
99
,
311
325
.
Kornberg
,
T.
(
1981
).
engrailed’. A gene controlling compartment and segment formation in Drosophila
.
Proc. natn. Acad. Sci. U.S.A
.
78
,
1095
1099
.
Kornberg
,
T.
,
Siden
,
L
,
O’farrell
,
P.
&
Simon
,
M.
(
1985
).
The engrailed locus of Drosophila’, in situ localization of transcripts reveals compartment specific expression
.
Cell
40
,
45
53
.
Levi
,
G.
,
Crossin
,
K. L.
&
Edelman
,
G.
(
1987
).
Expression sequences and distribution of two primary cell adhesion molecules during embryonic development of Xenopus laevis
.
J. Cell Biol
.
105
,
2359
2371
.
Levine
,
M.
&
Hoey
,
T.
(
1988
).
Homeobox proteins as sequencespecific transcription factors
.
Cell
55
,
537
540
.
London
,
C.
,
Akers
,
R.
&
Philips
,
C.
(
1988
).
Expression of Epi-1, an epidermis-specific marker in Xenopus laevis embryos, is specified prior to gastrulation
.
Devi Biol
.
129
,
380
389
.
Malacinsh
,
G. M.
,
Brothers
,
A. J.
&
Chung
,
H. M.
(
1975
).
Destruction of components of the neural induction system of the amphibian egg with ultra-violet irradiation
.
Devi Biol
.
56
,
24
39
.
Mangold
,
O.
(
1933
).
Uber die Induktionsfahigkeit der verschiedenen Bezirk der Neurula von Urodelen. Naturwissenschaften
21
,
761
766
.
Nieuwkoop
,
P.
&
Faber
,
J.
(
1967
).
Normal Table of Xenopus laevis (Daudin)
.
North-Holland: Amsterdam
.
Patel
,
N. H.
,
Martin-Blanco
,
E.
,
Coleman
,
K. G.
,
Poole
,
S. J.
,
Ellis
,
M. C.
,
Kornberg
,
T. B.
&
Goodman
,
S. G.
(
1989
).
Expression of engrailed proteins in Arthropods, Annelids and chordates (manuscript submitted)
.
Richter
,
K.
,
Grunz
,
H.
&
Dawid
,
I. B.
(
1988
).
Gene expression in the embryonic nervous system of Xenopus laevis
.
Proc. natn. Acad. Sci. U.S.A
.
85
,
8086
8090
.
Scharf
,
S. R.
&
Gerhart
,
J. C.
(
1980
).
Determination of the dorsal-ventral axis in eggs of Xenopus laevis: complete rescue of uv-impaired eggs by oblique orientation before first cleavage
.
Devi Biol
.
79
,
181
198
.
Scharf
,
S. R.
&
Gerhart
,
J. C.
(
1983
).
Axis determination in eggs of Xenopus laevis: a critical period before first cleavage, identified by the common effects of cold, pressure and ultra violet irradiation
.
Devi Biol
.
99
,
75
87
.
Sharpe
,
C. R.
(
1988
).
Development expression of a neurofilament-M and two vimentin-Iike genes in Xenopus laevis
.
Development
103
,
269
277
.
Sharpe
,
C. R.
,
Fritz
,
A.
,
De Robertis
,
E. M.
&
Gurdon
,
J. B.
(
1987
).
A homeobox-containing marker of posterior neural differentiation shows the importance of predetermination in neural induction
.
Cell
50
,
749
758
.
Spemann
,
H.
&
Mangold
,
H.
(
1924
).
Induction of embryonic primordia by implantation of organizers from a different species
.
In Foundation of Experimental Embryology (B. H. Willier & J. M. Oppenheimer, Eds)
, pp.
144
184
.
Hafner
:
New York
.